Research And Application Of Information Filtering Method

With the development of computer science and telecom technology, it becomes more and more easily to get messages. The spring up of Internet made the world enter the era of information. In the face of the information blast, how to find the useful and wholesome information, and avoid the useless and harmful information, which is always an issue deserved to study. Actually, only the useful information is what we need, much more useless or harmful information should be filtered.Most of the information filtering systems existed is key words based or rules based. There are also a few content-based systems. After studying and evaluating kinds of algorithms used in information filtering systems, two solutions are presented. These two solutions, combined with Nature Language Processing (NLP), adopt content-based KNN algorithm and Naive Bayes algorithm respectively. KNN algorithm is used in illegal web page filtering system and Naive Bayes algorithm, modified, is used in junk mail filtering system. These two algorithms are implemented on Linux operation system.These two applications are evaluated using international evaluation method, and the consequence shows that they both work well.